Connect with us

Noticias

Author Alok Sama on OpenAI, Bubbles, and When to Sell

Published

on

We’re talking tech.

Alok Sama is the former president and CFO of SoftBank Group International and the author of the new book The Money Trap: Lost Illusions Inside the Tech Bubble. In this podcast, Motley Fool host Ricky Mulvey caught up with Sama to discuss:

  • OpenAI’s latest fundraising round.
  • “Happiness for everyone” as an investment philosophy.
  • The illusion — and reality — of power.

To catch full episodes of all The Motley Fool’s free podcasts, check out our podcast center. To get started investing, check out our beginner’s guide to investing in stocks. A full transcript follows the video.

This video was recorded on Oct. 12, 2024.

Alok Sama: People often ask me you’re cynical about Silicon Valley, you’re cynical about Wall Street, but you seem to put Masa Son on a pedestal and that’s the reason. He had this quality that completely eludes me, and when he tells me that you’re not enough of a hunter, I get where he’s coming from. I don’t see what he does. I don’t see risk the way he sees it.

Mary Long: I’m Mary Long and that’s Alok Sama, former President and CFO at SoftBank International, and author of the new book, The Money Trap. Lost Illusions inside the Tech Bubble. SoftBank is a Japanese holding company that’s made big bets on some of the biggest names in Tech. Alibaba, ByteDance, Arm Holdings, also Uber, DoorDash, Lemonade, Opendoor, and a lot more. Some of those bets have paid off handsomely. Others quite the opposite. From 2014-2019, Sama was a top executive at SoftBank, where he worked closely with the firm’s founder, Masayoshi Son and had a front row seat as the company launched its legendary $100 billion vision fund. My colleague, Ricky Mulvey recently caught up with Sama for a conversation about the delicious circularity in AI markets. The difference between being a hunter and being a cook, what it means to live in the future, and knowing when to sell. Quick programming note before we get to today’s show. Since this conversation runs a bit longer than usual, we won’t be posting an episode tomorrow. Enjoy the day off Fools. We’ll see you again on Monday.

Ricky Mulvey: We’ll get into the book, but I want to start with this OpenAI news, because you’re comfortable talking about your former employer. You just wrote a book about them, and they’re back in the news with this investment in OpenAI. The nonprofit just raised $6.6 billion at a $157 billion valuation. SoftBank putting $500 million into the mix. What do you make of this valuation, this rise of OpenAI right now?

Alok Sama: Candidly, I think SoftBank investing 500 million is probably the least interesting aspect of the deal. You got to keep in mind as mind boggling as these numbers seem. SoftBank has when I last l close to 40 billion in cash, son signed his last AGM, talked about everything else he’s done previously is a warm up bag. You can put that together in the context of a man who’s been known for big, bold, audacious usually spectacularly successful moves. This is not his big moves, I guess. Let’s talk about OpenAI. To me, the question I get asked a lot is AI, is it overhyped? Are some of these valuations over hyped? When I see Safe Superintelligence, which is another company that raised pre-revenue. Forget about business model, just the concept from our OpenAI employee raising $1 billion. Is it a $5 billion valuation? OpenAI at 157 billion for a company that is bleeding cash, $5 billion from reports and beating people too. They lost their CTO and a whole bunch of people. You have to ask the question. A things getting overdone? Is there element of formal going on? I suspect in the private space there is because, unlike Google, for example, in its early days. with with OpenAI there’s real competition out there.

There’s Google for one, and getting AI right is existential for Google. Talk about deep pockets, Google’s got their own Gemini, LLM. You’ve got Facebook, again, with that monumental user base across its various platforms into Facebook Whatsapp, and they’ve got an open-source AI LLM out there. this competition, there’s real alternatives. Having said that, there’s no doubt, OpenAI is very real, first mover. Chat GPT has become almost synonymous with AI in the eyes of the world. There are strong traction in the enterprise user base. I I don’t argue with the fact that it’s worth a ton. I just wonder about a 150 billion number as I do about 5 billion for Safe Superintelligence for a company, that’s just belly on a concept. That to me and it’s a good thing this stuff is going on in the private markets because I’ve lived through inflated valuations 2000. You had these sorts of things going on in the public market at that point in time.

There are other aspects of OpenAI. Sorry, I’m branding on a little bit, I just find this so interesting that. The other thing that’s really interesting about OpenAI, is you’ve got five capital leading this round. They led the previous round, so they get to mark up the deal in the books at 2X. That’s unusual in VC Lane. People don’t very often do that. You’ve also got investors to the investors who are participating are Microsoft and NVIDIA. Guess what? OpenAI does with the money they raise. They take the cash, and they buy capacity on Microsoft Cloud. The money goes to Microsoft. What does Microsoft do with the money? Well, it buys NVIDIA chips. The money goes to NVIDIA. There’s almost a delicious sort of circularity. It’s a little bit incestuous too. There’s a lot going on in that deal and as great a company as it is, and Sam Ortman, terrific entrepreneur and obviously, a fundraiser and a salesman, too. I think there’s a real risk when you start extrapolating that valuation as people often do in terms of other things going on in the private and the public space.

Ricky Mulvey: What do you think that valuation is based on then? That $157 billion?

Alok Sama: I think that’s my point. I don’t know. Having said that, I’ve been in tech Investment land, and frequently, in my book, I tell the story of how Son, Sam and I have dinner with Mark Zuckerberg, and the one thing that came through to me loud and clear, is Masa’s regret, even pay a not investing in Facebook at a $10 billion valuation and I dare say when Facebook was served up at 10 billion, you and I might have had the same discussion or not come up with an answer, which is what is that 10 billion based on? Look that’s frequently the case with technology. Relatively early stage technology investments. It’s just very difficult to hang your hat on a specific valuation number. There is a huge leap of faith involved, and you take the view that look, there’s an asymmetric payoff. this thing from 150, it could be worth over 2 trillion the way NVIDIA is or Facebook is, right? That’s the only sensible response I can give you. There is no conventional valuation metric that would, as I sit here today, give me enough time to make something up, but I can’t come up with something on the spot of the moment.

Ricky Mulvey: In your book, you talked about these early stage companies, where they don’t have cash flow and that’s often that is what you value companies on, so you have to move to something like, what would it be gross merchandise value or the total amount of sales going on in an online marketplace and not necessarily the cut or the profit even that a company is being taken from that. We’re seeing a lot for the retail investors listening. What what are the company metrics that sort of set off your BS meter for these young emerging exciting companies?

Alok Sama: I think too much of an emphasis on revenue multiples, for example, you talked about GMV. There was a real heart button. GMV is gross merchandise value and, in the context of an e commerce company. you buy something on Amazon for 100 bucks. That’s the gross merchant value. Amazon’s cut of that might be, five or 10%. That’s Amazon’s revenue. You could have a huge GMV and not make any money at all. the 10% is revenue, and then there’s cost. When I see people extrapolating of revenue multiples into infinity, and the profit is a complete leap of faith, I start scratching my head. That to me is is a bit of a red flag. I don’t call it a BS. I think that’s way too strong. I think OpenAI is very real. It’s going to be worth a lot. I just don’t know how much, and I don’t know how to value it.

Ricky Mulvey: You’ve been a part of multiple bull cycles, and your book covers being a part of those, and people go with the herd. They react as if I’m not going to quote directly, but people react as if they’re on drugs when they’re part of these mad herds running toward these evaluations, and for us listening, you’d expect you’d expect the bankers to react as you right? Like, Mr. Spock evaluating these investing decisions with very cold, hard analytical skills. That doesn’t seem to be the case. It seems that emotions continue to rule no matter where you are in the investing spectrum.

Alok Sama: I think that’s always going to be the case. I think 300 years. Book I talked about this wonderful example of this book written by the Scotsman Charles Mackay. I think it was written in the 18th century and this guys like isit’s considered to be, like the ultimate authority on bubbles. He talked about. People might have heard of tulips in Meerland, and how ridiculous that became right there. But but this guy missed the biggest bubble in his own lifetime, which was railroads in England. Those of you might have watched Dan and abbey. Lord Grantham, that’s how he lost his shirt on [laughs] Railroad. This has been documented for a long time and I got to tell you I’ve been in and out of these markets for lost at this point, man, I’m showing my age almost 40 years. I get caught up, too. Everyone goes free.

But I tell you, the one thing that is it’s good, and this is actually really useful to talk about is to the extent there’s frat, a lot of that is in the private markets and not the public markets. The big difference relative to the Internet bubble is the hype at the time was in public markets. You had companies pre-revenue, going public at ridiculous valuation IPOs that were double 2x3x. I think what’s happened now, it’s interesting is you had so much access in the public markets, and we forgotten very quickly in 2021. 2021, you had I think the number is 1,100 IPOs, off which, 600 were SPACs. Remember those? Many of those just blew up. Some of those were like we work, which went to zero, others, 23 and me, for example. All these sacks went public at 10 bucks, but 23 and M is trading at pennies. Again, look at previous cycles, 2000, for example. It takes the IPO market at a minimum of four years to recover, and you haven’t had excess in the public markets. Yes, NVIDIA has been a rocket ship, but if you look at NVIDIA, it creates that depending on what your projection is between 35-40 times earnings. That’s not for a company with 75% profit margins and growing the way it has. That’s not outrageous. Outrageous was Cisco at 200 times earnings in year 2000. by the way, if you’d invested in Cisco in the year 2000, you’d still be losing. It’s the private markets that’s why there’s a lot of raised eyebrows.

Ricky Mulvey: Think why is there so much laggy you think in the private markets compared to the public markets right now, especially with Froth? You mentioned the spack lesson, where I got burned a little bit on a couple of SPACs that I got excited about that still haven’t recovered. But within the private markets especially, it seems to be ruled by almost the lesson that you mentioned earlier with Masa Son not investing in Facebook at $10 million. A lot of these private market investors, VCs, almost don’t lament the things they spent money on and then lost. They lament the opportunities that passed them by that ended up 100 X, 1,000 Xing, that thing.

Alok Sama: Look, these are some of the smartest, savviest people around, and that put Masa Keno Hill in terms of his ability to spot macro trends, and there’s no doubt that what’s going on with AI certainly at the enterprise level, where it’s proving out very quickly, truly is transformational as a productivity true. I think the jury is out in terms of at the individual level, how it might impact our lives. I haven’t seen any apps, the way we did with Web 2.0. Uber, etc. Maybe those will come, but the jury is still out. But at the enterprise level, it is transformational as a productivity true, and people get that, and there’s just not a lot of plays around. As a VC, it’s not a trend you can sit on out. It out. You’re seeing that. There are others, for example, I’m associated with Warburg Pincuse, which is not a household name, but one for people in the business would recognize it as absolutely cutting edge when it comes to growth investing, particularly in technology. They’ve chosen to set out these big flashy evaluations, and I focus instead on AI, the enterprise level and buying it to businesses where you could make a huge impact on margins, can top line, as well as bottom line, leveraging AI. it is I’m being up if I’m answering your question, but from a VC growth investive perspective, it’s like, you can’t afford to sit this out.

Ricky Mulvey: It’s something where with your analytical brain, you can’t value. If we’re moving toward the singularity.

Ricky Mulvey: How do you even put a price tag on that thing? Let’s get in.

Alok Sama: I do put a price tag on dreams a little bit. I may I use that expression in my book. It’s tough. The excitement is very real. I’m excited about it. I just can’t figure out or value it.

Ricky Mulvey: Let’s get into your journey at SoftBank. The investing style really revolves around the happiness for everyone pitch, this dream of Masa Son. While I know you’ve heard it a bunch of times, many of our listeners haven’t. What is happiness for everyone?

Alok Sama: What he’s trying to get at is idealism when it comes to technology investing. Let’s go back. General Electric used to be a technology company once upon a time. It’s Thomas Edison’s Company. You could make a legitimate case that electricity, light bubs, for example, contributed to helps for everyone. I think that’s a lot of what Son Sann’s talking about, which is leveraging technology to improve lives. Happiness, he can get overly philosophical about that, but that’s what he’s trying to get at. The first time I met him, I remember right in the lobby of his building, and then he talked about it when I met with him. He had this empathetic robot, Pepper, which was supposed to be a companion for the agent. This is again, AI powered with an emotional engine, supposed to be a companion in Japan, with an aging population. It’s a real issue with the highest suicide rates in the world. That’s addressing, that’s laudable. That’s a lot of where Masa is coming from. Very different from history of AI is a benign. It’s going to make lives better view, different from Elon who thinks that this might be something we have to run away from and colonize mass.

Ricky Mulvey: There’s also I was talking to Nate Silver on the show a few weeks ago. It’s the AI makes the world a worse version of a casino. You mentioned to the robot, and many of the predictions of Masa son came true. With the Pepper robot, you now have AI chat bots that are exactly like the movie Her, where people are having quote, real relationships with them in order to combat loneliness. That’s a whole rabbit hole, we have very complicated feelings. But for the base of this, it’s something that it came true. When you were going through these deals, was that something where, you would present a evaluation case, but you would also have to present this is how it makes the world happier with the companies we’re interested in investing in or did that come upstream before it got to you?

Alok Sama: Short answer is no. We didn’t have, I never thought about it, but, something like a happiness index or anything associated with an investment memo. But the thing about Masa, and this is what really got me hooked. This is a man who describes himself as a crazy guy who lives in the future. A lot of people think of investing as bits. Almost gambling. Son is not a man, you’ll find within a million mile radius of Vegas. She’s not a gambling man. I associate gambling with addictive thrill seeking behavior. That’s not him. A man who lives in the future and when you live in the future, you say tomorrow’s headlines, your assessment of risk is fundamentally different. It was often felt when I had doubts, reservations, who am I to hold him back? He backed Alibaba. Based on his vision for what e-commerce might unfold in China, based on looking at how it had evolved in the US. But I tell you, what’s even cooler? This is literally my favorite Masa Son story, and I think one that people will relate to.

I think we can all agree. We all felt that excitement when we held a smartphone and iPhone for the first time, I certainly did in my hands, which is, man, this is seriously cool. This is life changing. Now, but Masa Son, before the iPhone existed, as is talking about 2005, draws up a picture of something that looks like an iPhone and takes it to Steve Jobs and said, Steve, you’re the only one who can come up with something like this and when you do, you got to give me an exclusive for Japan. Steve tells him, you’re a crazy guy. I like you. I don’t have a product. You don’t have a phone company. What am I going to give you an exclusive for, but I like you. If I do something like this, I’ll give you an exclusive. On the back of this somewhat casual conversation, handshake, no contractual obligation. He goes and spends $20 billion buying or close to 20 billion buying this company in Japan, which is just bleeding cash. It is such a dog that Vodafone, which owned the company actually lends Masa money to take it off its hands. Steve Jobs comes up with the iPhone in early 2007, and gives Masa the exclusive, because he always saw this smartphones, he builds this company around marketing data oriented network, selling smartphones to the Japanese, and ends up on a mark to market basis when the company went public, making $40 billion. It’s a monumental number. Hundred billion mark to market on Arm, 75 billion on Alibaba, 40 billion on SoftBank, Japan, the mobile company. These are spectacular successes.

Ricky Mulvey: When you meet him through your friend, it was Nikesh. Where you met Masa Son and you said, I made Masa feel I understood his brand of genius. I’ll quote.” As a banker you have to make a lot of really wealthy people feel smart and understood”. This is that game at the highest level. How do you do that?

Alok Sama: I talk a little bit about that. I think my experience with CEOs is they like to talk about their business, their vision, and they don’t like being sold to. When you start reading from a pitch book, and you get into salesy mode, you denigrate yourself a little bit. My approach was always to look them in the eye, and as if I’m having a cup of coffee with them, having a drink with them, and let them talk and hopefully try and respond intelligently to what they say. You get up rapper going, and they come away saying that, OK, this is a guy I can relate to. He gets me, and when they have an idea or a problem they want to solve, hopefully they call you.

Ricky Mulvey: With Masa, you don’t need to convince him that, you’re also a crazy guy. Like he’s not looking for that? He’s not looking for a little bit of crazy.

Alok Sama: Well, look. One of the pivotal chapters in the book is I talked about this idea of hunters and cooks. Son Sann tells me that, you need to be more of a hunter. You’re basically suggesting I was a cook. Sorry, Wait.

Ricky Mulvey: You’re Michelin Cook.

Alok Sama: Well, that was a Michelin star chef man. That’s what I was fishing for, because I was good at what I did, which is getting deals done. No, but honestly, my role there, and I think that’s true for a lot of people around Son Sann is they tend to be facilitators facilitating his vision because the quality he has, living in the future, vision, as a futurist as a technologist, is just not something you can replicate. I did not have that. That is part of why people often ask me, you’re cynical about Silicon Valley, you’re cynical about Wall Street, which you seem to put Masa Son on a pedestal. That’s the reason. He had this quality that completely eludes me. When he tells me that you’re not enough of a hunter. I get where he’s coming from. I don’t see what he does. I don’t see risk the way he sees it.

Ricky Mulvey: One of the central themes of your book is the idea of power, and how there are people who you have an immense amount of power you’re flying on private jets. You’re moving in billions of dollars. Then there’s also times within your interactions with Masa Son, where you appear almost, like, more powerless. I would say one of those is within the WeWork deal where Softbank goes on to invest tens of billions of dollars in WeWork. You seem to be in the spot where you’re like, this person is a silver tongued Adam Neumann, like a silver-tongued snake oil salesperson. Feel free to disagree with my assessment. I see a shake of the head.

Alok Sama: Wow, two things. I talk about power, but not to be overly philosophical about this, but I talk about power being an illusion. In the epilogue, book starts with a dramatic scene where I’m at Claridges Hotel in Mayfair in London with two most agents, if you can believe. They tell me I have power because I control billiger dollars, and my feedback to them is power. That’s an illusion. I can barely control my own bladder. It gets philosophical when I lose both my parents, my dog within a very short period of time. It just brings home that the notion that you control anything in your life is a bit of an illusion. That’s a bit of a philosophical as opposed to an investment point. In the context of WeWork, look, it’s Adam Neumann it’s a cool fun story. He’s been a movie made about it. It’s a fun movie. I enjoyed it. Jared Leto does a great job. Son, he acknowledged it himself. I think his exact words were. He said that I blame myself even more than Adam Neumann. That’s the nature of the VC game. You’re not gonna get everything. You just got to keep in mind that he might have lost $10 billion on WeWork, which is monumental by any standards, but he also made 100 billion on Arm. I take that trade any day.

Ricky Mulvey: Well, I’m going to stand on the philosophical point for a sec, because I think it’s interesting where you put that. There’s multiple definitions of power, and we almost need different words for it. There’s the idea that you cannot control almost the acts of uncontrollable natural forces, sickness, that thing. We have that word for power, but then we also have the ability to make people do things. That is also combined by power and when billions of dollars is being put into WeWork, when billions of dollars is being put into chip companies, there’s a business model that you’re also a part of that has led to things like people getting free Uber rides. That I’ve been on the downstream of SoftBanks actions in the market. That is an intense form of power, man. That’s also very real. It’s different from the other thing you’re describing.

Alok Sama: Look, that’s fair. I think the impact that SoftBank had on the investment landscape in the Valley is profound. I talk the book about how Sequoia, which is King of the Hill in the Valley. That’s Google, you name it. Every major technology success story they’ve been behind. I’m a bit of a golf nut and I described the meeting with Doug Leone, who’s the managing partner, Sequoia. His reaction using a golfing metaphor. I go back to Bobby Jones, arguably the greatest of all times when he sees Jack Nicklaus play and his reaction is, he plays a game with which I’m not familiar. What Masa did with the vision Fund did was a game changer and just to be very specific, Sequoia, billion dollar funds were rare in the value. Sequoia was probably the biggest one. I think that fund was two billion. The next fund turned out to be eight billion.

Everyone felt they needed a bigger boat. That when that $100 billion fund came along. I think the total volume of VC investing in the previous year was 70 billion. This was monumental. We talked about OpenAI toward the beginning of that discussion. It’s these large mega-cap technology funds, venture slash growth funds, with a ton of liquidity and that’s a little bit of what you’re saying. They need to deploy a lot of money in the new thing. You still saying that. A little bit of that is the SoftBank effect. A little bit of that is also this weird notion we had of modern monetary theory and printing money. It’s a combination of factors.

Ricky Mulvey: Sequoia, the relationship with them wasn’t always rosy. They were upset at SoftBank because of essentially the valuations and the money that the investment fund was putting in to smaller companies to which they were also invested, which on the surface, you would think, great, more money going into an investment I own, the valuations getting larger. I’m making a lot of money off of this, but that wasn’t the case always with Sequoia.

Alok Sama: No. I think their objection to be fair was, again, it comes back to the notion of power and control. The conventional VC approach is you sit on the board and you drift feed capital. Like you’re talking about tens of millions at the time, I gave you a certain benchmarks, I evaluate you against that, you hit your goals, I’ll give you no money, so on so forth. I very tightly controlled allocation capital. Nothing wrong with that. It serve people like Sequoia brilliantly. Masa’s view is, I used the expression.

He treated founders, the way they want to be treated by an investor, which is I love you, man. It is I give the example of Radesh, which is what came up in the context of Sequoia. Radesh got married, Masa flew from Japan to India to attend the guys’ marriage. It’s that personal level of engagement that belief in an entrepreneur. I believe in you, I back you, whatever capital you need. I’m behind you. I think that’s a little bit of what the VCs were struggling with. That’s one of the things that’s very different about how Masa approach to investing.

Ricky Mulvey: One of the things you write about, let’s talk about selling. This is something a lot of our listeners think about, especially, we’re heavy into a mega trend. We’re back into a bull market. You right, “While Masa had an impressive track record investing ahead of technology mega trends, he frequently held on to long.” We’ll start with what led you to that conclusion, and then I’ll ask you what you learned from it. Why do you think Masa held on too long? Because frequently we hear the reverse that investors are too tradery, and they sell too much, and that hurts them.

Alok Sama: I think let’s look at that comment was made at probably halfway through the book and just based on the observation, that South Banks Yahoo stake, Yahoo was king of the hell. Some of us who we are old enough to remember, like late ’90s, 2000, it was the portal to the web. His stake in Yahoo at its peak was 30 billion. He ended up selling it at seven billion. I think the same thing turned out to be true with Alibaba. I think on a mark to market basis at one point, Masa that stake would have been worth, I want to say 200 billion or something very close to that. At its peak, he ended up liquidating it at 75 billion. Now, it is still a spectacular success story, or those are still spectacular success stories. But obviously, had he sold off at a different time, he would have made a lot more, which would suggest that, he’s not a great trader, but man that’s a completely different skill set. That’s what prompted that remark.

Ricky Mulvey: The other side of this though is Nvidia. When SoftBank didn’t have an investment, and, it’s better to have loved than who have never loved at all. But what led to the sale in Nvidia? Because this is what happened to the fact.

Alok Sama: That’s a great question. I think we can learn something from that. First of all, two things about Nvidia. Early 2017, Masa wanted to buy the company Nvidia. With a controlled premium, we could have bought the company for $80 billion. Could have been a history, it’s worth over two trillion now. We all know that. CFIs, which is the Committee for Foreign Investment to the United States, there is no way, the US government would have CFIs. Acting through CFIs would have allowed a deal like that, foreign ownership of something as strategic to AIs and Nvidia, so that was off the table. What Masa ended up doing is buying in the fund. When you buy on your own balance sheet, if you’ve actually bought a company, the way he did on, with the intention of holding on, making some operational changes, as opposed to buying on a leverage basis in a fund with this pressure to recycle fund to investors, your mindset changes a little bit, and that’s what happened with Nvidia in the fund, which is, he on a leverage basis, made a brilliant return. But if he’d held on, that that 4x, 5x could have been 20x.

Ricky Mulvey: Some of that, did that have to do with the funds deal where it was paying out, what was it like a 7% dividend to all of its holders through out which is uncommon for a tech when we think of high tech investments, you’re going for those big home runs. You’re not going for these steady dividend payments.

Alok Sama: The division fund was completely unique, in as much as there was a layer of leverage provided by the limited partners. That was part of the fund. The investors, the limited partners, mainly the Middle Eastern investors out of Saudi Arabia and the United Arab Emirates, they also provided capital in the form of mezzanine, which paid a cash dividend 7%. The fund had to pay them cash, and that created pressure to realize liquidity and that again, is at least a small part of the explanation, that when you have an opportunity to liquidate something, you try and do it.

Ricky Mulvey: You told the story earlier of the dinner you had where you were described as a cook and not a hunter. Your job changed dramatically at SoftBank, around 2019. You went from president and CFO to senior advisor. When did you feel your job really changing? That we talked about power as a nebulous thing. When did you feel your power softening and feeling almost sideline moved out of one of the top seats there?

Alok Sama: I think that happened earlier. When I left and became a senior advisor, I was basically out. That’s the way I wanted it, and that’s the way we negotiated. Someone wanted me out of the vision fund. When I was shut out of the vision fund, even though there was some very interesting stuff that was going on outside of the vision fund, for example, SoftBank owned Sprint and I let the merger of Sprint with T mobile. There was a lot to do outside of the vision fund, but I was excluded from the vision fund, and at the time, SoftBank focus was very much on the vision fund. That was an element, I think it’s fair to say of being if not sideline, certainly being one step removed from where the real action was. That’s where it started to happen in terms of the drift.

Ricky Mulvey: Did a two part question, then I went down a digression, I’m going to get back to the two part question now that I remember it. What did you learn about when to sell from your experience at SoftBank? How does that affect you as an investor now?

Alok Sama: My personal philosophy is, that whole line about, bulls make money, bears make money, pigs get slaughtered. My personal view of the world is, man, if I run into a situation where I’ve doubled my money or tripled my money or whatever, I’m out of there. I tend not to get too greedy. But that’s not in the world of technology, that is most certainly not the right answer. Because if you’ve got a winner in the world of technology, you need to be thinking, not just 10x, but, like, potentially a 100x. Like Masa Son, you look at the Alibaba success story, that’s like, 40 million, eventually, 75 billion realized. Now, it would be really unfair to criticize him for not realizing 200 billion because it is arguably the most spectacular investment all time. My mindset is a little bit conservative because, I’m particularly at my stage in life. I’m thinking in terms of my nest egg for retirement. But if you want to be a successful technology investor, you’ve got to think in terms of really backing your winners. You’re seeing that with the funds. Sequoia started this is frequently because of the venture funds. Because of pressure to recycle capital, they were getting out too early, they created a separate fund where you could hang on much much longer. It’s like, Google computer next google. You bought Google at IPO, it’d be a spectacular investment.

Ricky Mulvey: The money trap lost illusions inside the tech bubble. I want to mention this to listeners. You didn’t get a ghost writer for this. It is beautifully, right?

Alok Sama: What? No. Quite the contrary, I did not set off to write this book, but I wanted to be a writer, and I spent two-and-a-half years in graduate school. I got an MFA at New York University. Just in experimenting, like all students do, was like, I was writing and throwing stuff up on the wall and see what’s Dick’s thing. That’s how this book came about. It’s been a real labor of love for me.

Ricky Mulvey: It’s a lyrical, it’s beautifully written. I really enjoyed the stories in it. I learned some stuff for my investing brain. I think our listeners will get a lot out of it. There’s a lot of big business people when they write a memoir, sometimes it’s a history written things that’s given off to a ghost writer. Your book is the complete opposite of that. It’s thoughtful, it’s warm, it’s personal. I really enjoyed it, and I’m grateful for the time and insight.

Alok Sama: Thanks, Wicky.

Mary Long: As always, people on the program may have interest in the stocks they talk about, and the Motley Fool may have formal recommendations for or against, so don’t buy yourself stocks based solely on what you hear. I’m Mary Long. Thanks for listening. We’ll see on Monday Fools.

Continue Reading
Click to comment

Leave a Reply

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *

Noticias

Estoy usando chatgpt, copilot y gemini para encontrar nuevas aplicaciones de Android

Published

on

Todavía encuentro que la incorporación de IA en mi trabajo diario es ligeramente incómodo, incluso si tiene algunos casos de uso probados que lo hacen útil. Para poner esto a prueba, decidí usar AI para solucionar un problema subyacente que tengo en la tienda Google Play: utilicé Google Gemini, Microsoft Copilot y ChatGPT para ayudarme a descubrir nuevas aplicaciones. De esa manera, puedo afeitarme el tiempo que paso explorando Reddit y X para nuevas aplicaciones de Android para probar mi confiable Samsung.

Experimenté usando Gemini, Copilot y Chatgpt. Tenía curiosidad sobre cuál de los tres me ofrecería más variedad en opciones y si estos chatbots reconocerían solicitudes específicas como “aplicaciones gratuitas solamente”. Aquí están los resultados que encontré.

Relacionado

Google ha hecho que buscar en la tienda de Play sea tan frustrante que ya ni siquiera me molesto

Arregla tu tienda, Google

Enumeré las recomendaciones de la aplicación para cada chatbot de IA e incluí el número de descarga en la tienda Google Play entre paréntesis para medir la popularidad. Mi esperanza es descubrir nuevas aplicaciones que no necesariamente aparezcan en los resultados de búsqueda inmediatos de Play Store/menos populares.

Experimento 1: Encontrar nuevas aplicaciones de actualización del clima

Aviso utilizado:

Hi [AI]! I'd like to find a new Android app that can tell me the weekly and daily weather forecast. Please give me free apps only.

ChatGPT recomienda:

  • Accuweather (100m+)

  • El canal meteorológico (100m+)

  • Clima subterráneo (10m+)

  • Viento (10m+)

  • Google Weather (1+)

  • 1weather (100m+)

Copilot recomienda:

  • 1weather (100m+)

  • FlowX (500k+)

  • El canal meteorológico (100m+)

  • Accuweather (100m+)

  • Clima impresionante – Yowindow (10m+)

Géminis recomienda:

  • Accuweather (100m+)

  • El canal meteorológico (100m+)

  • Weathercan (500k+)

Comparando los tres modelos, ChatGPT salió en la cima cuantitativamente. Chatgpt me dio seis recomendaciones, Copilot tenía cinco, y Gemini solo dio tres. Otra nota es que me gustó que Gemini considerara mi ubicación (incluso si es un poco espeluznante) proporcionar una recomendación más relevante. De lo contrario, la lista de aplicaciones de los tres era bastante segura.

El resultado final

ChatGPT reconoció qué aplicaciones meteorológicas eran gratuitas o gratuitas con soporte de anuncios. Pero también dio una lista genérica de las mejores aplicaciones del clima, que no quería. Por otro lado, Copilot especificó si estas aplicaciones eran gratuitas en las dos primeras entradas pero luego se detuvieron. Sin embargo, proporcionó un abastecimiento, por lo que podría confirmar el contexto (a diferencia de los otros dos chatbots). Géminis me dio una recomendación interesante adaptada a mi ubicación (Weathercan). Pero tampoco especificó ningún precio de aplicación. En general, la experiencia estuvo bien. Todavía prefiero la selección de Copilot debido a que 3/5 es nuevo.

Ejemplo de copiloto respondiendo a mi consulta para encontrar nuevas aplicaciones meteorológicas y mostrar fuentes proporcionadas

Experimento 2: Encontrar nuevas aplicaciones de toma de notas

Aviso utilizado:

Hi [AI], I am in need of a new note-taking app. I would prefer if it included a place to import PDF documents and had a handwriting mode. I would also like an option for online storage syncing.

ChatGPT recomienda:

  • Notabilidad (no en Play Store)

  • Goodnotes (1m+)

  • Microsoft OneNote (500m+)

  • Evernote (100m+)

  • Notebook Zoho (5m+)

  • Xodo (10m+)

Copilot recomienda:

  • Goodnotes (1m+)

  • Notabilidad (no en Play Store)

  • Microsoft OneNote (500m+)

  • Evernote (100m+)

  • Tejido líquido (no en Play Store)

Géminis recomienda:

  • Goodnotes (1m+)

  • Notabilidad (no en Play Store)

  • Microsoft OneNote (500m+)

  • Nebo (500k+)

  • Samsung Notes (1B+)

Si revisamos mi mensaje, notará que cometí un error y olvidé especificar la plataforma. Como resultado, estaba preocupado de recibir aplicaciones de toma de notas solo en iOS.

Ejemplo de las recomendaciones de aplicaciones de toma de notas de Gemini que toman notas por plataforma

Sin embargo, dadas las respuestas, me di cuenta de que los chatbots eran en su mayoría perdonando esa omisión. Por ejemplo, Gemini reconoció que no especificé la plataforma, pero aún rompí las recomendaciones por plataforma.

Relacionado

Gemini podría responder pronto todas las consultas de sus aplicaciones sin salir de Play Store

‘Ask Play About esta aplicación’ está en pruebas

El resultado final

Me impresionó la respuesta de Gemini a mi solicitud. Me dio una buena lista de aplicaciones (menos notabilidad) y la clasificó por plataforma. También reconoce la anotación de PDF y la escritura a mano. Los listados de copilotes eran más genéricos y no especificaron las plataformas como lo hizo Gemini. Tampoco proporcionó ningún abastecimiento (como lo había hecho en la solicitud anterior).

Chatgpt salió en la cima; No solo reconoció todas mis especificaciones (opciones de sincronización, escritura a mano y soporte para PDF), sino que también las enumeró punto por punto e incluía las plataformas disponibles. Además, descubrí aplicaciones de las que ni siquiera había oído hablar (Zoho Notebook y Xodo). En general, la experiencia fue adecuada, considerando mi aviso inicial.

Experimento 3: Recomendaciones de juego

Como beneficio adicional, intenté obtener recomendaciones de juegos para novelas visuales basadas en misterio y juegos de rompecabezas similares a Danganronpa. Desafortunadamente, los resultados fueron increíblemente decepcionantes, así que aquí hay un desglose de lo que ocurrió.

Aviso utilizado:

​​​​Hi [AI], I am looking for paid Visual Novel and puzzle game recommendations on the Play Store that match the style and themes of Danganronpa.

ChatGPT recomienda:

  • The Arcana: un romance místico (1m+)

  • Ace Abogado Trilogía Phoenix Wright (10k+)

  • Synchronicity de Dead: Mañana llega hoy (no en Play Store)

  • The Nonor Games: 999 y la última recompensa de Virtue (no en Play Store)

  • ¡Doki Doki Literature Club! Puerto móvil (no en Play Store)

  • Elección de juegos: elección de robots (10k+)

  • Reigns: Su Majestad (100k+)

  • La edad silenciosa (no en Play Store)

  • Profesor Layton y Curious Village Via Emulator (no en Play Store)

  • Planeta destrozado (no en Play Store)

Copilot recomienda:

  • Cero Escape: The Nonor Games (no en Play Store)

  • Trilogía de abogados de as (10k+)

  • La casa en Fata Morgana (no en Play Store)

  • Steins; puerta (10k+, requiere crunchyroll)

  • Muerte de la muerte (no en Play Store)

Géminis recomienda:

  • Puertos móviles de la serie Danganronpa (1K-10K)

  • Serie de abogados de as (10k+)

  • ¡7days!: Mystery Visual Novel (5m+)

  • ARGO’s Choice: Visual Novel (100k+)

No me gustó la mayoría de estas recomendaciones. ChatGPT proporcionó una lista extensa, pero la mayoría no estaba en Play Store. El copiloto no pudo recomendar los títulos de Play Store (solo 2/5). Gemini desperdició una recomendación sobre el título real en lugar de un juego similar similar, además de 2/4 fueron títulos gratuitos, que no es lo que quería.

Por otro lado, tanto ChatGPT como Copilot siguieron las instrucciones para exhibir títulos pagados. También esperaba ver a la tribu nueve en estos listados, un nuevo título de Gacha de los creadores de Danganronpa; Se ajustaría a la factura de los temas, pero es un servicio en vivo, lo que ya significa que es un no-go basado en los criterios que proporcioné. Debería omitir a Tribe nueve según mis especificaciones, y lo hizo.

El contexto es clave al conversar con AI

A partir de estos experimentos, dominar sus indicaciones es más importante que hacer la pregunta. De lo contrario, la IA solo sugerirá las aplicaciones más populares con un desglose general de lo que hacen, lo que no es útil. Si tiene necesidades específicas, debe incluirlas en sus indicaciones/indicaciones de seguimiento. No incluí una plataforma en mi segundo intento y recibí algunas recomendaciones que no me aplicaron. En contraste, utilicé un ejemplo específico en mi experimento de juego, lo que lleva a malos resultados. Para aplicaciones generalizadas, dio recomendaciones sólidas. La clave está en algún punto intermedio.

Mi única preocupación es cómo maneja nuevas aplicaciones. Los chatbots públicos se ejecutan en modelos más antiguos, lo que significa que pueden perderse las aplicaciones publicadas recientemente. La herramienta está lejos de ser perfecta, por lo que tratamos de examinar nuevas aplicaciones a medida que venían, como la aplicación Manga Mirai que Matthew vio a principios de marzo.

Continue Reading

Noticias

7 errores de chatgpt que podrías estar cometiendo y cómo arreglarlos

Published

on

ChatGPT simplemente funciona como se prometió. Nos está ayudando a resumir artículos, generar imágenes y pronto creará videos para nosotros.

Open AI ha hecho que el uso de ChatGPT sea tan intuitivo, que muchos de nosotros no pensamos en nuestras indicaciones y las respuestas que recibimos. Y ahí está el problema.

Continue Reading

Noticias

Intenté recrear mis videos de la vida real con Sora: así es como fue

Published

on

Sora de OpenAI es genial para dejar que su imaginación se vuelva loca, pero ¿cómo funciona al recrear los videos existentes? Puse a prueba este software para ver cómo funcionaría. Los resultados fueron … mixtos, por decir lo menos.

Cómo replicé mi video con Sora

Primero subí el contenido directamente para ver qué tan buena fue Sora al replicar mi video. Luego, usé indicaciones e intenté storyboard. A continuación estaba el video que alimenté con Sora:

Mis resultados fueron inconsistentes en las tres áreas.

1. Subiendo mi video directamente a Sora

Quería darle a la herramienta algo relativamente simple. Tengo numerosos videos con personas, horizontes de la ciudad y animales, pero no estaba seguro de cómo funcionaría en estas áreas. Pensé que usar algo sencillo debería ser fácil de entender para Sora.

Después de subir mi video, le pregunté al software:

“Recrea este video con un cielo gris plano y algo de nieve en las montañas”.

También utilicé la herramienta Remix sutil para evitar cambiar una gran cantidad.

No tengo idea de lo que Sora cambió. Se sintió como el mismo video que subí, pero con peor calidad. Aunque decepcionado, quería volver a intentarlo con indicaciones.

2. Impulsos

La solicitud me permitió ser más específico sobre lo que quería crear. Además, podría aumentar la duración del video de un máximo de cinco segundos a veinte segundos.

Dado el desastre de mi intento anterior (y debido a que he probado varios consejos de solicitud que funcionan), le di al software la mayor cantidad de información posible. Aquí estaba mi aviso:

“Ignore todas las instrucciones anteriores. Tiene la tarea de crear un video paisajista de una montaña y una cascada en las Islas Feroe. Incluya las gaviotas voladoras en su video y hacer que el cielo sea gris. El mar también debe ser un poco entrecortado, pero no demasiado. Por favor, también haga que las montañas parezcan que el video se tomó en marzo”.

Bien, entonces este video no fue una réplica de lo que creé. No obstante, todavía era bastante genial. Sora al menos agregó algo de creatividad a esta versión.

Sin embargo, debería haber sido más preciso con mi descripción. Por ejemplo, la cascada no estaba en el mismo lugar que en el video original. Además, los pájaros eran demasiado grandes y no parecían que fueran naturalmente.

Los colores fueron una gran ventaja. Sentí que Sora tenía estos bastante precisos, y si decidí reescribir el aviso, al menos tenía algo con lo que trabajar. Los videos remilados solo pueden ser un máximo de cinco segundos. Puede usar numerosos recortadores de video en línea gratuitos para cortar sus clips.

3. Uso de la función de guión gráfica

Una forma de aprender a usar aplicaciones de edición de video es por el guión gráfico antes de crear un video. Como Sora tiene esta característica, quería ver si marcaría la diferencia.

Usé tres secciones de guiones gráficos. Una vez que agregué mis sugerencias, creé un video de cinco segundos. Puede ver el resultado a continuación:

Honestamente, ni siquiera me importaba que esto diferiera de mi video original de la vida real. Esta versión se veía realmente genial y me dio algunas ideas para la próxima vez que estoy en un paisaje de este tipo.

Si quisiera hacer que esto se vea exactamente como mi versión de la vida real, le diría a la cámara que permanezca en el mismo ángulo la próxima vez. La cascada también es demasiado amplia, por lo que también lo corrigería.

¿Con qué funcionó Sora bien?

Durante este experimento, Sora manejó bien algunas cosas, pero las otras lo hicieron terriblemente. Esto es lo que me gustó de la herramienta.

1. Una buena función de guión gráfica

Mi video favorito de los tres intentos fue el que creé con mi guión gráfico. Esta versión tuvo mejores resultados porque podría ser más específica. Además, la herramienta sabía exactamente dónde incluir cada elemento.

Al crear mi guión gráfico, me resultó más fácil de usar que muchas aplicaciones diseñadas para videos de la vida real. Todo fue intuitivo y receptivo, lo que ayudó masivamente.

2. Variando ángulos de cámara

Si bien quería que Sora se quedara con un ángulo de cámara, me gustó descubrir que podría usar diferentes para mis videos. Las imágenes donde la cámara voló cerca de la cascada era particularmente fresca.

En el futuro, usaré diferentes ángulos de cámara y otros consejos útiles de Sora para mejorar mis videos.

¿Dónde podría haber mejorado Sora?

Puedo ver el potencial de Sora, pero fue decepcionante cuando recreé mis videos. La aplicación necesita arreglar tres elementos antes de que me sienta cómodo vuelva a ejecutar este experimento y obtener mejores resultados.

1. Edición de video más precisa

Sora no parece manejar muy bien la edición de video. Cuando subí mis propias imágenes, todo lo que recibí a cambio era una versión de peor calidad de lo mismo. Quizás mis indicaciones debían ser más precisas, pero también sentí que el software jugaba un papel aquí.

En lugar de solicitar, creo que tener botones como la extracción de fondo funcionaría mejor.

2. Significaciones de video más largas

Estoy seguro de que Sora me permitirá hacer videos más largos en el futuro, pero subir contenido preexistente durante un máximo de cinco segundos fue frustrante. Este no es tiempo suficiente para ser verdaderamente creativo.

Si bien el límite de 20 segundos en los videos que creo en la aplicación es mejor, todavía es a veces limitante. Supongo que crear múltiples videoclips y reunirlos en una aplicación de edición de video externa. Por ejemplo, podría usar una de las alternativas a Capcut.

3. Mejores animaciones para personas y animales

Sora parecía funcionar bien con los paisajes, pero no se podía decir lo mismo de los animales. Por ejemplo, los pájaros volando en mis videos parecían muy antinaturales. En lugar de ir a algún lado, estas aves estaban efectivamente de pie en el aire.

Otros también se han quejado de lo mala que es Sora en las interacciones de los objetos. Me imagino que el software planchará esto a medida que obtenga más información y, con suerte, lo hace en poco tiempo.

¿Qué tipo de videos funcionan mejor con Sora?

No recomiendo usar Sora para recrear videos de la vida real. Si bien podría haber hecho ciertas cosas de manera diferente, el software no me impresionó.

En cambio, creo que Sora es mejor para crear videos desde cero. Ofrece muchas opciones si desea dejar que su creatividad funcione salvaje con indicaciones y guiones gráficos. Del mismo modo, usaría la herramienta para inspirarse en futuros proyectos de la vida real.

Continue Reading

Trending